search

firasm / CEST

Analysis of studies
2 stars 0 forks source link

Create function to correct for B0 shift #29

Closed firasm closed 9 years ago

firasm commented 9 years ago

I need to create a function that fits just the water peak in the CEST spectrum pixel by pixel and shift the spectrum accordingly.

The new spectrum and x-values should be saved as adata.

This is from Piotr's suggestion to average neighbouring voxels to try and increase SNR.

firasm commented 9 years ago

This function is now coded and seems to work.

Here is a 0 sat freq image for a mouse brain on the left, and the corresponding Water Shift Map.

unknown-4

Here are the two plots for the shifted and non-shifted CEST spectra for two neighbouring pixels and their means.

unknown unknown-2

I'm going to close this now, and task @srveale with testing it in phantoms in different situations in issue #31

DrSAR commented 9 years ago

In the image on the left ("CEST image at SatFreq=0ppm") above, is this an image after or before the shifting of spectra happened? In other words, is this an image of the minimal intensity over the entire z spectrum for the entire image? If so, I'm surprised that there is this one strong deviation near the star-labelled symbol. Water shift map looks good. This will prove useful in cleaning up data...

firasm commented 9 years ago

It's an image before the spectra shift.

Basically, the upshot of this is that the Water shift map looks similar to the "CEST image at SatFreq=0ppm" after shimming. I showed this in issue #26, reproduced here for convenience:

7c87cd76-d592-11e4-85df-bd43914c7197